The introduction of insulin provided an effective form of therapy and it was thought that, with careful treatment, diabetics would have relatively normal longetivity. By the 1940s, it had become apparent that the life expectancy and the quality of life in insulin dependent diabetics was diminished by what may be called the late complications or sequelae of diabetes: retinopathy, nephropathy, neuropathy, premature peripheral and coronary artery disease. Evidence from laboratory, experimental animal and human studies suggests that the late complications are related to the abnormal metabolism of diabetes. But, proof that aggressive diabetic treatment and resultant strict control of the abnormal metabolism results in decreased late complications is not available. A large scale, prospective, radomized therapeutic trial employing current methodology should provide an answer to this question. In this application we propose to: 1) determine whether experimental therapy with home glucose monitoring and insulin infusion pumps of randomly assigned insulin dependent diabetics results in glucose control which is a significant improvement over that achieved with conventional therapy, 2) determine the extent and nature of problems or adverse effects which may result from such experimental therapy--a. metabolic b. neuropsychological c. psychosocial, 3) determine whether measurements of red blood cell sorbitol and/or glycosylated proteins besides hemoglobin provide improved assessment of diabetic control, 4) demonstrate our ability to quantify retinopathy, peripheral nerve function, and arterial hemodynamics in insulin dependent diabetics.